BY Fast Company 4 MINUTE READ

WWDC 24, as it’s known, is arguably more critical for Apple’s future than any other WWDC in recent memory. That’s because Apple is widely expected to focus on its artificial intelligence plans, specifically the AI features it will integrate across its major operating systems such as iOS 18 for iPhone, and macOS 15 for its Mac computers, when the software ships this fall.

Apple is widely considered to be late to the AI game, which kicked off with the release of ChatGPT in late 2022. Since then, virtually every major tech company, including Google, Microsoft, Samsung, and Amazon has gone all-in on AI. But Apple has been slow to adopt the technology, much to the annoyance of its consumers and investors.

The company is expected to add dozens of new AI features across its operating systems on Monday. But Apple has much bigger and more important questions to address regarding its AI plans. Those include:

WHO ARE APPLE’S AI PARTNERS?

There’s no doubt that Apple is playing catch-up when it comes to operating in the AI space. Due to that, the company is expected to rely on third-party AI partners to help power some of the AI features it will roll out in the beta versions of its new software on Monday. The most important question that Apple needs to answer, then, is: Who are its AI partners that will power its new AI features?

Rumors have suggested that Apple is close to or has already inked a deal with Open AI, the company behind ChatGPT. But Apple has also been reported to be in talks with Google regarding its Gemini large language model (LLM).

Both partners would have their advantages and disadvantages for Apple. If Apple chose Open AI, it would get access to what is widely considered the most advanced LLM technology on the planet and offer it to its millions of iPhone users. However, Open AI is a controversial company, and there are questions surrounding the legality of some of its training data. Open AI’s CEO, Sam Altman, is also a controversial figure who seems to have as many detractors as admirers in the AI space.

Because of this, a partnership with Google for its Gemini LLM technology may be the safer bet for Apple. But Google’s Gemini is not as advanced as Open AI’s ChatGPT. Worse, the company’s new AI search answers feature has given the public the impression that the company’s AI just isn’t that accurate or trustworthy.

Still, Google and Apple have worked together since the launch of the iPhone back in 2007. Google Maps was the default mapping service on the iPhone until 2012, and Google Search has been the default search engine on iPhone since 2007. Though they are also competitors, Google and Apple make good partners and it may be easier for the two companies to work together than Open AI and Apple.

Apple will also need to address who its AI partner is in China, which will likely be Baidu. China is a very important market for Apple, and the Chinese government requires that any AI models be vetted by its internet regulator, the Cyberspace Administration of China, meaning Apple will likely have to use a Chinese LLM partner for iOS 18 in the country.

HOW WILL APPLE HANDLE AI PRIVACY?

The next question Apple needs to answer is how it will handle privacy regarding its new AI features. This is incredibly important, because AI tools aren’t known for privacy. The only way they can work so well is by sucking up a huge amount of data to train on—and if you want that data to be personalized to an individual, it needs access to that individual’s data.

And let’s just be frank: Google, Open AI, and Baidu are not known for their privacy protections, even though Apple is. That means that Apple will need to reassure users that, no matter whom it chooses as its AI partner, their data will stay as secure as possible.

How Apple can do this technically remains to be seen. The company will probably use on-device processing for many of the iPhone’s more basic AI features, like generative AI photo editing and summarizing emails and articles on the web. But when it comes to an AI-powered Siri, which may use Google or Open AI’s online servers to process queries, Apple needs to explain how it’s planning to keep user queries and data safe from those data-hungry companies.

WHICH DEVICES WILL SUPPORT APPLE’S NEW AI, BOTH ON-DEVICE AND IN THE CLOUD?

Make no mistake: Apple may have dragged its feet getting into the AI race, but the company is well-positioned to become the largest AI company by number of users overnight when it releases iOS 18, iPadOS 18, watchOS 11, and macOS 15 this fall.

Apple has billions of devices in use around the world (back in November, OpenAI said it had around 100 million weekly ChatGPT users). But just because Apple has billions of active devices in use around the world doesn’t mean that all of them will run its latest AI features.

In order for a device to run AI it needs to have a relatively modern processor and plenty of RAM. These hardware requirements, however, can be skirted if the AI is sent to more remote servers to be processed—but that presents privacy issues.

So Apple needs to make clear which of its devices can run its new AI features on-device, resulting in the most privacy, and which devices will need to rely on remote servers to run its new AI features.

On-device AI processing will likely require the latest iPhone 15 Pro and M3 and M4 Macs. If this is the case, Apple will need to make clear to users with devices lacking the appropriate hardware that their AI queries will need to be processed remotely.

Apple will make its AI announcements—and hopefully answer these three important AI questions—during its Worldwide Developers Conference keynote on Monday, June 10, 2024.

FastCompany